L.E.D. headlights on Locomotives

[M636C]

in order to conserve power (and some electrical 'driver' expense) some LED installations 'multiplex' power to the individual diodes, much as older digit displays did to individual numbers or segments.  (Quickly move an older alarm-clock display in the dark in your peripheral vision to see the effect.)  This is what I think produces the 'disappearing headlight' effect on the ACS64s and other contemporary locomotives.  It is of very little consequence to human vision, as the persistence of vision is far longer than CCD 'shutter' times, but it is probably important in the design of things like outward-facing cameras that are expected to work effectively.

I've taken hundreds of photographs of Pacific National NR class locomotives fitted with LED headlights and they have never appeared to be off. In fact, I have been able to count the number of individual diodes in each "lamp", these being arranged as a regular hexagon just smaller than the circle of the AAR standard headlight mounting. Of the top of my head I recall the number as 64 but I may be wrong.

I assumed the "disappearing headlight" reference was to the ditch lights. Ours flash alternately when the horn is sounded and since drivers normally sound the horn (and get a response) from a photographer, I usually have one of the two ditch lights dark.

The ditch lights are quite different from the headlights, consisting of a circle of maybe eight individual LEDs, each looking like a small bulb. Again these are seen as all on or all off in photos with shutter speeds up to 1/2000th.

Marker lights are different again, generally consisting of a small number of small LED units like those in the headlights, maybe eight individual LEDs. These can usually display either red or white (presumably using different LEDs)

When viewing an oncoming locomotive, the LED headlight appears to cause more glare than a sealed beam although the effect in photographs is not as pronounced as to the naked eye.

So my assumption is that the ACS64 has different headlights to the Cv40-9i

Peter

[blue streak 1]

Before further discussion of LED headlights.  Are they powered by battery or from an AC source ?  LEDs by their nature are DC powered.  Most home LEDs use a blocking diode that only allows for a half wave of the AC power to go thru the diode to activate the LED.  Any one have an idea of how long after after power shut off to a LED does it continue emitting light ?  More than 1/60th of second ?

From what has been posted here about photos seems to indicate that headlight LEDs are being powered by battery ?   

[Anonymous]

blue streak 1

Before further discussion of LED headlights.  Are they powered by battery or from an AC source ?  LEDs by their nature are DC powered.  Most home LEDs use a blocking diode that only allows for a half wave of the AC power to go thru the diode to activate the LED.  Any one have an idea of how long after after power shut off to a LED does it continue emitting light ?  More than 1/60th of second ?

From what has been posted here about photos seems to indicate that headlight LEDs are being powered by battery ?   

 

I don't see any reason for LED headlights being battery powered. Allmost all locomotives have at least an inermediate DC circuit. Diesel-electrics produce AC current that is rectified to DC for use in DC traction motors. For AC traction motors the intermediate DC current is converted to AC of variable frequency.

On electrics the AC current is provided by the catenary. The locomotive internals are similar.

I looked for LED headlight and found these: https://www.jwspeaker.com/products/led-headlight-model-8770s-locomotive/

http://www.hydra-tech.net/wp-content/uploads/2016/05/locomotive-Spec-Sheet-May-June-.pdf

https://www.railheadcorp.com/product/ke-par56-75v-led/

The voltage for these is 32 VDC or 75 VDC. There should not be a problem to provide these voltages from the intermediate DC circuit and smooth the voltage for LED use.
Regards, Volker
Regards, Volker

[Paul of Covington]

   Unlike incandescent lights where brightness varies with the voltage applied, LEDs are constant voltage devices, and brightness is controlled by regulating the current flowing through them.   While it is true that LEDs are DC devices, as I understand it, rather than applying a steady current to them, it's common practice to apply a series of high frequency pulses to the LED.   This way they can drive it with a higher peak current producing a brighter light than would be possible with steady DC.   The average current thus stays below the amount that would burn up the LED.

[Paul of Covington]

blue streak 1

Any one have an idea of how long after after power shut off to a LED does it continue emitting light ?

   I'm not sure, but I think it's pretty close to instantaneous.   It's the persistence of vision in the eye that makes it appear steady, as with fluorescents.

[Overmod]

Note that the LED itself is the “blocking diode’ that converts AC to half-wave DC (with corresponding effective ‘ripple’). The issue is in providing the appropriate constant-current power control to keep the voltage across the diode strictly limited under all conditions — and it likes to run away as the device heats.

A main purpose of phosphors in LEDs is to extend the effective ‘on’ time to reduce flicker when driven from pulsed drivers.  This is similar to the use of long-persistence phosphors in older CRTs displaying ASCII data with the kind of long refresh interval encountered in the ‘olden days’.

The ‘off time’ with an LED is determined by the time needed for the electrons and holes to finish recombining across the band gap and the last of the photons emitted in that process to be generated.  That is not a long time, and the physical area involved is usually very small, so the device itself goes out promptly when current is not present

[erikem]

IIRC, the main use of the phophors in "white" LEDs is to downconvert the blue light of the driving LED to lower frequency light (i.e. red to green), with reduing flicker a secondary role (think fluorescent lamps with a visible light driver instead of UV from the mercury arc). The phosphors lose at least 15% or so of the blue light photon energy in the downconversion process, leaving a "hole" light specrum in the upper green to lower blue range. Color rendition could presumably be improved by placing some bare ~590nm LED's in the mix to help fill in that gap.

My understanding is that the turn-off time for bare LEDs is in the nanosecond range.

[rdamon]

LEDs like 1.8-3.3VDC ..  Beyond a voltage drop resistor there is a capacitor in the AC->DC circuit. The drain on the capacitor is why it looks like the light turns off slow.